Embodiments of the invention include systems, methods, and computer-program products for lifting metadata from financial documents to allow for automated exception processing. As such, allowing for automated decisions for exception processing to systematically be resolved base on matches between lifted metadata. The exceptions may include one or more irregularities such as bad micro line reads, outdated check stork, or misrepresentative checks that may result in a failure to match the check to an associated account for processing. As such, once an exception is identified during the processing the metadata lifted from the document with the exception may be utilized to search financial records at the financial institution to attempt to identify the correct version of the document. Subsequently, the system may correct the irregularity systematically and automatically.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A system for exception processing of financial documents, the system comprising: a memory device with computer-readable program code stored thereon; a communication device; a processing device operatively coupled to the memory device and the communication device, wherein the processing device is configured to execute the computer-readable program code to: receive an indication of a user transaction, wherein receiving an indication of the user transaction comprises receiving a copy of one or more financial documents associated with the user transaction for processing; identify data from the financial document, wherein the data from the financial document is extracted from the financial document using optical character recognition and the data is stored as metadata; process the financial document by identifying exceptions in the metadata for the financial document being processed, wherein identifying exceptions further comprises identifying a type of financial document as a credit document or a debit document; determine known data associated with financial accounts of a user based on the metadata, wherein the financial accounts are identified by comparing the metadata that is not associated with the exception to the financial institution records to generate the known data; compare the exceptions identified in the metadata associated with the financial document to the known data associated with the financial accounts of the user; match an appropriate financial account to the financial document comprising exceptions, wherein the match is based on the comparison between the extracted metadata from the financial document comprising exceptions to the known data associated with financial accounts; identify one or more resolutions associated with the identified exception, wherein identifying one or more resolutions further comprises comparing the extracted metadata with data associated with either financial institution records or financial accounts of the user and correcting, automatically, the identified exceptions in the financial documents based on the one or more resolutions.
A system for automatically processing financial documents with errors uses a computer to: receive digital copies of financial documents like checks, identify data from these documents (payor, payee, amount, etc.) using optical character recognition (OCR) and store it as metadata; detect exceptions or irregularities in the metadata (e.g., bad micro line read); determine known user account data by comparing the error-free metadata with financial institution records; compare the identified exceptions to the known account data; match the financial document to the correct account based on this comparison; identify potential automatic corrections by comparing the extracted metadata with institution records or user account details; and automatically correct the errors in the financial documents. The system identifies the type of financial document as credit or debit.
2. The system of claim 1 , wherein processing the financial documents further comprises directing a financial document without an exception to an appropriate financial account associated with the financial document such that payment for the user transaction associated with the financial document is completed.
In addition to the automated exception processing described previously, financial documents without errors (identified in the initial metadata extraction and validation step) are automatically routed to the appropriate financial account, completing the user transaction associated with the document. This means that if the system receives a check image and all the data is correctly read and validated, the system automatically initiates the payment process without requiring manual intervention.
3. The system of claim 1 , wherein comparing the extracted metadata from the financial document comprising exceptions to known data associated with financial accounts further comprises: compiling metadata associated with complete portions of the financial document comprising exceptions.
When comparing metadata from a financial document with errors to known financial account data to find a match, the system compiles the metadata associated with the complete, error-free portions of the financial document. This involves extracting all the data that *can* be reliably read from the document despite the presence of errors elsewhere. For example, if a check has a bad micro line but the payee and amount are clearly readable, the system compiles the payee and amount as known metadata.
4. The system of claim 3 , wherein comparing the extracted metadata from the financial document comprising exceptions to known data associated with financial accounts further comprises: compiling metadata associated with partially extracted portions or exception portion from the financial document comprising exceptions.
Building upon the previous claim of using complete portions of a document with errors, the system also compiles metadata from partially extracted or error-containing portions of the financial document. This means even if some parts of a data field are unreadable or questionable, the system attempts to capture whatever information it can. For instance, if only part of the routing number on a check is readable due to a smudge, the system will save the readable digits for matching.
5. The system of claim 4 , wherein comparing the extracted metadata from the financial document comprising exceptions to known data associated with financial accounts further comprises: comparing both the metadata associated with complete portions of the financial document and metadata associated with partially extracted portions of the financial document to known data associated with financial accounts.
Combining the previous two descriptions (complete and partial data extraction), the system compares *both* the metadata extracted from complete portions *and* the metadata from partially extracted portions of the financial document containing errors to the known data associated with financial accounts. This increases the chances of finding a match even when data is incomplete or potentially inaccurate. The matching process considers all available clues from the document.
6. The system of claim 5 , wherein comparing the extracted metadata from the financial document comprising exceptions to known data associated with financial accounts further comprises: determining a match between a known financial account and the financial document comprising exceptions based on the comparison of both complete and partial portions of the financial document comprising exceptions.
Using both complete and partial data from error-containing financial documents to find a match, as described previously, the system determines a match between a known financial account and the document based on comparing both the complete and partial portions of the extracted metadata. The system uses a combined analysis of all available data to increase the likelihood of correctly identifying the intended account, even with incomplete or corrupted information.
7. The system of claim 1 , wherein identifying data from the transaction documents associated with the user transaction further comprises capturing via optical character recognition, images of various elements of the transaction documents, wherein the various elements are unique to the transaction documents, wherein the identified data is stored as metadata.
To identify data from transaction documents, the system captures images of various elements of the documents using optical character recognition (OCR). The elements captured are unique to the transaction documents being processed. The identified data is stored as metadata for later use. The elements being captured, like payee name, are unique and distinct on the financial document.
8. The system of claim 7 , wherein elements of the financial document that comprise the identified data includes one or more of a payor name, a payee name, date, payment amount, account number, or routing number.
As part of identifying data on financial documents via OCR, as described previously, the specific elements that are captured and stored as metadata include one or more of the following: payor name, payee name, date, payment amount, account number, or routing number. These are considered key data points for matching documents to accounts.
9. The system of claim 1 , wherein the financial documents comprise one or more of a check, deposit ticket, automated teller machine (ATM) receipt, or general ledger ticket.
The financial documents processed by the system for automated exception handling, as described previously, can include one or more of the following: a check, a deposit ticket, an automated teller machine (ATM) receipt, or a general ledger ticket. The system is designed to handle various types of financial documents.
10. The system of claim 1 , wherein exceptions comprise one or more irregularities in the financial documents such as bad micro line reads, outdated check stork, or misrepresentative checks.
The "exceptions" identified by the system during financial document processing, as described previously, can include one or more irregularities such as bad micro line reads on checks, outdated check stock, or misrepresentative checks. These exceptions prevent normal, automated processing.
11. The system of claim 1 , wherein correcting, automatically, the identified exceptions in the financial documents further comprises: determining a confidence rating for each of the one or more resolutions, based on comparing the extracted metadata with the data associated with financial institution records; and transmitting an authorization request to an external system, based on determining that the confidence rating associated with the one or more exceptions is below a predetermined threshold.
When automatically correcting identified exceptions in financial documents, as described previously, the system determines a confidence rating for each potential correction based on comparing the extracted metadata with financial institution records. If the confidence rating is below a set threshold, the system transmits an authorization request to an external system (presumably for human review or further validation) before applying the correction.
12. The system of claim 1 , wherein metadata comprises both structural and descriptive metadata, wherein the structural and descriptive metadata includes design, specification, and details from the financial document relating to one or more of account data, dates, payee, payor, addresses, routing numbers, or payment amounts.
The metadata used by the system for exception processing includes both structural and descriptive information derived from the financial document. This structural and descriptive metadata contains designs, specifications, and details from the financial document related to one or more of account data, dates, payee, payor, addresses, routing numbers, or payment amounts.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
February 29, 2016
May 2, 2017
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.